System and method for intelligent analysis of full life cycle requirements

Abstract

The present application provides a full life cycle demand intelligent analysis system and method, relates to the product life cycle management technical field, and the present application constructs a unified data set and extracts standardized demand characteristic parameters by acquiring multi-source demand data of each stage of the product life cycle; a demand evolution index is constructed to realize dynamic quantitative evaluation of the demand state, and a threshold value is determined to trigger a graded demand adjustment strategy; a design consistency index and a demand stability index are constructed to realize two-dimensional analysis of system consistency and stability, and a differentiated collaborative optimization strategy is triggered; a demand deviation propagation index is constructed to quantitatively analyze the deviation source and propagation path, and a closed-loop optimization control is realized in combination with a convergence determination mechanism. The present application can realize accurate identification and dynamic regulation and control of the demand evolution process, and improve the system operation stability and the intelligent level of demand management.

Description

Technical Field

[0001] This invention relates to the field of product lifecycle management technology, specifically to a full lifecycle demand intelligent analysis system and method. Background Technology

[0002] With the increasing complexity of products and the widespread application of multi-stage collaborative development models, the entire product lifecycle, from requirement identification and design implementation to production execution and operational feedback, exhibits highly dynamic and coupled characteristics. In practical applications, user requirements are often characterized by high uncertainty, frequent changes, and ambiguous expression, leading to continuous evolution of requirements at each stage of the lifecycle and a chain reaction affecting design parameters, production execution, and resource scheduling. Simultaneously, information transmission between different stages suffers from lag and distortion, causing requirement deviations to gradually accumulate and even amplify during transmission, ultimately triggering systemic deviation risks.

[0003] Existing technologies typically focus on single-stage requirement management or localized optimization, such as merely recording and version-controlling requirement changes, or independently monitoring and analyzing production execution deviations, lacking a unified modeling and collaborative analysis mechanism throughout the entire lifecycle. Especially during requirement evolution, there is a lack of effective quantification methods for the coupling relationships between multiple factors such as requirement conflicts, implementation deviations, and user feedback differences, making it difficult to accurately characterize the changing trends of requirement states and their impact on subsequent stages. Furthermore, existing methods generally fail to establish propagation path models for requirement deviations across different stages, making it impossible to effectively trace the sources of key deviations and identify high-impact stages, thus limiting the development of targeted control strategies.

[0004] On the other hand, at the optimization and control level, existing technologies mostly rely on static thresholds or empirical rules for adjustment, lacking data-driven dynamic feedback and adaptive optimization mechanisms. When the system experiences demand imbalances or deviation spreads, it can often only make local corrections, making it difficult to form a collaborative optimization strategy that links across stages. This leads to repeated problems, affecting the overall system's operational stability and product quality consistency. Summary of the Invention

[0005] The purpose of this invention is to provide an intelligent analysis system and method for full lifecycle demand in order to solve the problems mentioned in the background art.

[0006] To achieve the above objectives, the present invention provides the following technical solution: The whole lifecycle demand intelligent analysis method includes the following steps: S1. Parse and obtain multi-source demand data streams at each stage of the product lifecycle, extract demand semantic data, design execution data, production deviation data and operation and maintenance feedback data, and perform unified identification encoding, timestamp alignment and semantic normalization processing on various types of data to build a unified dataset. S2. Extract the frequency of requirement changes, degree of requirement conflict, degree of deviation of requirement implementation, deviation of user feedback, and duration of each phase based on the unified dataset, and perform dimensionless processing to establish a requirement feature set. S3. Based on the demand feature set, construct the demand evolution index, and filter according to the demand evolution index to determine whether the demand is in a stable, unbalanced or abnormally out-of-control state. Based on this, execute the hierarchical demand adjustment strategy, including demand freezing, priority adjustment and production correction, and form a demand adjustment execution dataset. S4. Based on the demand adjustment execution dataset, calculate the design consistency index. At the same time, calculate the standard deviation based on the stage sequence data of the demand evolution index and perform inverse mapping to obtain the demand stability index. Then, screen and determine the system consistency and stability status. Trigger differentiated collaborative optimization strategies according to different combination states, including backtracking correction, targeted compensation or stability enhancement adjustment. Record the adjustment results and construct a collaborative optimization execution result dataset. S5. Based on the collaborative optimization execution result dataset, calculate the demand deviation propagation index to conduct a quantitative analysis of the sources and propagation paths of demand deviations; determine whether the deviation is in a convergence state based on the demand deviation propagation index.

[0007] Furthermore, a unified dataset is constructed, including the following steps: Real-time monitoring of user interaction behavior at each stage of the product lifecycle; collection of textual requirements submitted by users, timestamps of requirement submission, and priority identifiers of requirements; and construction of user requirement input data. Real-time monitoring of parameter evolution behavior during product design process; collection of design version iteration count, design parameter modification records, and time information corresponding to each change; and construction of design parameter change data. The execution status of the production process is monitored in real time, and actual process operation parameters, design target parameters and the deviation between the two are collected. At the same time, production cycle time data is collected to construct production execution deviation data. Real-time monitoring of equipment status and user feedback behavior during product operation and maintenance; collection of equipment vibration amplitude, equipment temperature change, operating current fluctuation and speed change data; collection of equipment operation log records, fault alarm records and user satisfaction score data; and construction of operation and maintenance feedback data. User input data, design parameter change data, production execution deviation data, and operation and maintenance feedback data are processed in a unified manner. A unified identifier label is added to data from different sources through execution data identification and encoding methods. A timestamp alignment method is used to synchronize the time sequence of multi-source data. Semantic normalization processing is used to eliminate data expression differences and construct a unified dataset.

[0008] Furthermore, extract the frequency of requirement changes, the degree of requirement conflict, the deviation from requirement implementation, the deviation from user feedback, and the duration of each phase, including the following steps: Extract requirement identifiers and requirement submission timestamps from user requirement input data, and extract design parameter modification records and version iteration time information corresponding to the requirement identifiers from design parameter change data. Using a preset time window as the statistical unit, accumulate the number of change operations corresponding to the same requirement identifier and count the total number of requirement items within the current time window. Proportionate the number of change operations to the total number of requirement items to obtain the requirement change frequency. The textual requirement content and priority identification information are extracted from the user requirement input data, and the textual requirement content is semantically parsed to obtain the semantic expression result of the requirement. At the same time, the design parameter modification records in the design parameter change data are combined to identify the overlap of different requirement items in terms of resource occupation and functional realization goals. Based on semantic consistency analysis and resource conflict judgment rules, requirement items with logical conflicts or resource conflicts are marked and their number is counted. Then, the total number of requirements in the same stage is counted, and the number of conflicting requirement items is proportionalized to the total number of requirements to obtain the requirement conflict degree. The actual process operation parameter data collected by industrial sensors is extracted from the production execution deviation data, and the corresponding design target parameter data is extracted from the design parameter change data. Each set of actual process operation parameters and design target parameters are compared item by item to obtain the deviation range between each parameter. The deviation ranges of all parameters are summarized to form the overall deviation level, and the overall deviation level is used as the deviation degree of requirement realization. Extract user satisfaction rating data and user evaluation records within the corresponding time period from the operation and maintenance feedback data, obtain the pre-set target satisfaction reference value, perform difference analysis on the user satisfaction rating data and the target satisfaction reference value to form the deviation result, and use the deviation result as the user feedback deviation. The start and end timestamps corresponding to each lifecycle stage are extracted from user demand input data, design parameter change data, production execution deviation data and operation and maintenance feedback data. The time difference between the start and end timestamps is calculated to obtain the actual duration of each stage, and the actual duration is used as the stage duration. The maximum and minimum values ​​of each parameter in the historical data sample are obtained, and the frequency of requirement change, degree of requirement conflict, degree of requirement implementation deviation, user feedback deviation and stage duration are normalized based on the interval scaling method. The processing results are mapped to the [0,1] interval to obtain normalized requirement features, and a requirement feature set is constructed based on the normalized requirement features.

[0009] Furthermore, a demand evolution index is constructed, including the following steps: Extract the frequency of demand changes, degree of demand conflict, degree of demand implementation deviation, user feedback bias, and stage duration from the demand feature set, input them into the demand evolution modeling process, and perform weighted coupling calculation according to the historical regression weight coefficients corresponding to each parameter to obtain the demand evolution index.

[0010] Furthermore, triggering demand adjustment strategies includes the following steps: By comparing the demand evolution index with the preset first-level evolution threshold and second-level risk threshold, the demand is identified as being in a stable, unbalanced, or abnormally out-of-control state. When in an unbalanced state, a 10% to 20% freeze ratio is implemented for change requests corresponding to the frequency of demand changes, and a priority reordering adjustment of no less than 30% is implemented for conflicting requests corresponding to the degree of demand conflict. At the same time, a 5% to 10% reverse correction adjustment is implemented for production execution parameters corresponding to the degree of demand realization deviation to reduce the accumulation of deviations. When in an abnormal out-of-control state, a mandatory freeze of no less than 50% is implemented for change requests corresponding to the frequency of demand changes, and a priority reordering adjustment of no less than 60% is implemented for conflicting requests corresponding to the degree of demand conflict. At the same time, a mandatory reverse correction adjustment of 15% to 30% is implemented for production execution parameters corresponding to the degree of demand realization deviation, and cross-stage demand retrospective analysis is triggered simultaneously to eliminate systemic risks. The tiered demand adjustment strategy is transformed into a set of control instructions that include the demand freeze ratio, priority adjustment weight, and production parameter correction range, and a demand adjustment execution dataset is constructed; and distributed to the demand management, design adjustment, and production execution stages according to the product life cycle stage mapping relationship.

[0011] Furthermore, the demand consistency index and demand stability index are constructed, including the following steps: Based on the actual process operation parameters in the production execution deviation data and the design parameter modification records in the design parameter change data, the parameter identification alignment method is used to match the design parameters with the actual process operation parameters. The design target parameter value corresponding to the i-th parameter and the actual execution parameter value corresponding to the i-th parameter are extracted according to the unified parameter number order. At the same time, when there are performance indicators or result evaluation data in the operation and maintenance feedback data, the data mapping method is used to convert the performance indicators into corresponding parameter dimension data and supplement them into the actual execution parameter values. The design phase target values ​​and actual execution results are matched item by item, and the absolute value of the difference between each matching item is calculated. After the absolute value of all differences is averaged, the reverse normalization calculation is performed to obtain the design consistency index, which is used to characterize the degree of consistency between the design target and the actual execution results. Based on the output results of the demand evolution index in each stage of the product life cycle, the time series extraction method is used to obtain the demand evolution index sequence data corresponding to multiple stages. The mean calculation method is used to obtain the overall mean of the data in each stage of the sequence. Then, mean centering is performed on the data of each stage to obtain the deviation value. Finally, the cumulative square and square root calculation method is used to perform statistical processing on all deviation values ​​to obtain the standard deviation of the demand evolution index. The standard deviation of the demand evolution index is normalized using a reverse mapping method to obtain a demand stability index that characterizes the stability of fluctuations in the demand evolution process.

[0012] Furthermore, a dual-dimensional evaluation framework for consistency and stability, along with corresponding optimization strategies, is constructed, including the following steps: The design consistency index is compared with the preset design consistency judgment threshold, and the demand stability index is compared with the preset operation stability judgment threshold; thus, scheduling fluctuation risk and resource mismatch risk are screened out. To address the dual risks, a forced collaborative optimization strategy is triggered, which performs an overall backtracking correction on the design target parameters, implements a 20% to 30% regression correction on key execution parameters according to the deviation ratio, redistributes resource priority weights by 15% to 25%, and compresses or extends the production scheduling cycle by 10% to 20% to eliminate the risk of coupling between systemic deviations and fluctuations. To address the risk of resource mismatch, a targeted consistency correction strategy is triggered, which performs a fine-grained compensation adjustment of 10% to 20% on the parameters that cause the deviation, while maintaining the current scheduling rhythm to avoid introducing new fluctuation risks. To address scheduling fluctuation risks, a stability enhancement strategy is triggered, which smoothly adjusts the resource scheduling rhythm by 10% to 25%, while dynamically rebalancing the resource allocation ratio of high-fluctuation links by 10% to 20% to improve the continuity of system operation; and records the results of parameter correction, resource reallocation and scheduling adjustment to construct a collaborative optimization execution result dataset.

[0013] Furthermore, a demand deviation source tracing and impact path analysis is constructed, including the following steps: Based on production execution deviation data, design parameter change data, demand adjustment execution dataset, and collaborative optimization execution result dataset, this study uses a phase division and data association mapping method to identify and divide each stage of the product lifecycle. Actual execution deviation data for each stage is extracted from the production execution deviation data. Simultaneously, design target parameter data from the design parameter change data is combined to calculate and normalize the difference between the actual deviation and the design target deviation for each stage. The total deviation across the entire lifecycle is then proportionally allocated to obtain the deviation contribution value for the k-th stage. Based on the demand adjustment execution dataset and the collaborative optimization execution result dataset, a multiple regression analysis method is used to fit and calculate the impact of demand adjustment behavior and collaborative optimization adjustment behavior on the overall deviation propagation, obtaining the stage influence weight corresponding to the k-th stage. Finally, a weighted cumulative calculation method is used to sum the deviation contribution value and stage influence weight for each stage to construct a demand deviation propagation index.

[0014] Furthermore, an execution feedback and closed-loop optimization and update mechanism is constructed, including the following steps: By comparing the demand deviation propagation index with the preset optimization convergence judgment threshold, the risk of deviation propagation and diffusion and the risk of path amplification are screened out, triggering the iterative optimization strategy. The stage influence weight corresponding to the high-impact stage is suppressed by 5% to 15%, and the demand adjustment strategy and the collaborative optimization strategy are simultaneously implemented to strengthen the intervention and reduce the intensity of deviation propagation. The stage impact weights after strategy adjustment are fed back into the demand deviation propagation index calculation process, and the demand adjustment execution dataset and collaborative optimization execution result dataset are updated synchronously to form a full life cycle closed-loop optimization control mechanism driven by the demand deviation propagation index.

[0015] Furthermore, the full lifecycle demand intelligent analysis system includes: The full lifecycle requirement data fusion and processing module is used to parse and obtain multi-source requirement data streams at each stage of the product lifecycle, extract requirement semantic data, design execution data, production deviation data and operation and maintenance feedback data, and perform unified identification encoding, timestamp alignment and semantic normalization processing on various types of data to build a unified dataset. The standardized requirement feature parameter construction module is used to extract requirement change frequency, requirement conflict degree, requirement implementation deviation degree, user feedback deviation and stage duration based on a unified dataset, and perform dimensionless processing to establish a requirement feature set; The demand evolution assessment and control module constructs a demand evolution index based on the demand feature set, and filters according to the demand evolution index to determine whether the demand is in a stable, unbalanced or abnormally out-of-control state. Based on this, it executes a graded demand adjustment strategy, including demand freezing, priority adjustment and production correction, and forms a demand adjustment execution dataset. The consistency and stability analysis and optimization module is used to adjust the execution dataset according to demand, calculate the design consistency index, calculate the standard deviation based on the stage sequence data of the demand evolution index and perform inverse mapping to obtain the demand stability index, and then screen and determine the consistency and stability status of the system. It also triggers differentiated collaborative optimization strategies based on different combinations of states, including backtracking correction, targeted compensation or stability enhancement adjustment, while recording the adjustment results and constructing a collaborative optimization execution result dataset. The consistency and stability analysis and optimization module calculates the demand deviation propagation index based on the collaborative optimization execution result dataset, and performs quantitative analysis on the sources and propagation paths of demand deviations; it also determines whether the deviation is in a convergence state based on the demand deviation propagation index.

[0016] Compared with the prior art, the beneficial effects of the present invention are: This invention constructs a unified dataset by uniformly identifying, encoding, aligning timestamps, and semantically normalizing user requirement input data, design parameter change data, production execution deviation data, and operation and maintenance feedback data. Furthermore, it extracts parameters such as requirement change frequency, requirement conflict degree, requirement implementation deviation, user feedback deviation, and stage duration to establish a requirement feature set. Based on this, a requirement evolution index is constructed through weighted coupling calculation, enabling a quantitative description of the overall state of a product during multi-stage requirement changes. Compared to traditional analysis methods that rely on only a single stage or data source, this invention more comprehensively reflects the trend of requirement changes and their coupling relationships, thereby significantly improving the accuracy of requirement analysis and the reliability of decision-making basis.

[0017] This invention also introduces a design consistency index and a demand stability index to evaluate the system's execution status from two dimensions, and combines a demand deviation propagation index to quantitatively analyze the sources and propagation paths of deviations. Under a multi-threshold judgment mechanism, it triggers hierarchical demand adjustment strategies and collaborative optimization strategies, realizing coordinated control from demand adjustment and parameter correction to resource scheduling. At the same time, through dynamic adjustment of the stage influence weights and continuous feedback of strategy execution results, a closed-loop optimization control mechanism for the entire life cycle is formed. Compared with existing static rules or local optimization methods, it can effectively suppress deviation propagation and risk accumulation, improve system operation stability, and enhance the model's adaptive optimization capability to complex demand change environments. Attached Figure Description

[0018] Figure 1 This is a schematic diagram illustrating the overall execution of the method of the present invention; Figure 2 This is a schematic diagram of the overall method steps of the present invention; Figure 3 This is a schematic diagram of the overall system flow of the present invention. Detailed Implementation

[0019] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0020] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and those skilled in the art can make similar extensions without departing from the spirit of the invention. Therefore, the invention is not limited to the specific embodiments disclosed below.

[0021] Example 1 Please see Figures 1 to 3 This invention provides a technical solution: a method for intelligent analysis of demand throughout the entire lifecycle, the specific steps of which include: S1. Parse and obtain multi-source demand data streams at each stage of the product lifecycle, extract demand semantic data, design execution data, production deviation data and operation and maintenance feedback data, and perform unified identification encoding, timestamp alignment and semantic normalization processing on various types of data to build a unified dataset. S2. Extract the frequency of requirement changes, degree of requirement conflict, degree of deviation of requirement implementation, deviation of user feedback, and duration of each phase based on the unified dataset, and perform dimensionless processing to establish a requirement feature set. S3. Based on the demand feature set, construct the demand evolution index, and filter according to the demand evolution index to determine whether the demand is in a stable, unbalanced or abnormally out-of-control state. Based on this, execute the hierarchical demand adjustment strategy, including demand freezing, priority adjustment and production correction, and form a demand adjustment execution dataset. S4. Based on the demand adjustment execution dataset, calculate the design consistency index. At the same time, calculate the standard deviation based on the stage sequence data of the demand evolution index and perform inverse mapping to obtain the demand stability index. Then, screen and determine the system consistency and stability status. Trigger differentiated collaborative optimization strategies according to different combination states, including backtracking correction, targeted compensation or stability enhancement adjustment. Record the adjustment results and construct a collaborative optimization execution result dataset. S5. Based on the collaborative optimization execution result dataset, calculate the demand deviation propagation index to conduct a quantitative analysis of the sources and propagation paths of demand deviations; determine whether the deviation is in a convergence state based on the demand deviation propagation index.

[0022] In this embodiment, by constructing a demand data processing and analysis process covering the entire product lifecycle, and combining multi-indicator modeling, hierarchical adjustment, and closed-loop optimization mechanisms, dynamic perception, accurate judgment, and real-time intervention of demand status are achieved, effectively improving the systematicness, foresight, and overall optimization capabilities of demand management.

[0023] Figure 1This paper demonstrates the core technical path of the method, from multi-source data acquisition and feature parameter extraction to multi-level state evaluation and policy adjustment, ultimately achieving closed-loop optimization. The multi-source data unification and feature extraction flowchart corresponds to steps S1 and S2. Specifically, it expresses how the system standardizes and integrates requirement data from different sources and in different formats, and extracts key features such as requirement change frequency parameters, requirement conflict degree parameters, requirement implementation deviation parameters, user feedback deviation parameters, and stage duration parameters, thus establishing a requirement feature set.

[0024] The dynamic assessment and adjustment flowchart for demand evolution corresponds to step S3. Specifically, it describes how the system calculates the demand evolution index and compares it with preset first-level evolution thresholds and second-level risk thresholds to determine whether the demand is currently stable, unbalanced, or abnormally out of control. Based on this, it triggers hierarchical demand adjustment strategies such as demand freezing, priority adjustment, and production parameter correction.

[0025] The consistency and stability two-dimensional optimization flowchart corresponds to step S4. Specifically, it expresses that the system simultaneously calculates the design consistency index and the demand stability index, and compares them with their respective judgment thresholds to determine the consistency and stability status of the system, thereby triggering differentiated collaborative optimization strategies such as backtracking correction, targeted compensation, or stability enhancement adjustment.

[0026] The deviation propagation analysis and closed-loop control flowchart corresponds to step S5. Specifically, it describes how the system calculates the demand deviation propagation index, quantitatively analyzes the sources and propagation paths of demand deviations, and compares them with the optimization convergence judgment threshold. If convergence fails, the system iteratively strengthens and adjusts the weights of high-impact stages and related adjustment and optimization strategies, feeding the updated results back to the preceding calculation and control processes, thus forming a closed-loop optimization control mechanism throughout the entire lifecycle.

[0027] Example 2 Please see Figures 1 to 3 This embodiment, following the explanation in Embodiment 1, specifically describes the construction of a unified dataset, which includes the following steps: Real-time monitoring of demand interaction behavior at each stage of the product lifecycle; by using the log collection proxy module deployed on the server side of the demand management system to listen to the user demand submission interface; using interface call capture method and log parsing method to collect the text demand content, demand submission timestamp and demand priority identifier submitted by the user, and construct user demand input data, denoted as Du; Real-time monitoring of parameter evolution behavior during product design process is performed by embedding a version tracking and recording module in the product lifecycle management system (PLM). The version differential comparison method and parameter change tracking method are used to collect the number of design version iterations, design parameter modification records, and time information corresponding to each change, and construct design parameter change data, denoted as Dd. Real-time monitoring of the execution status during the manufacturing process is achieved by installing industrial sensors and data acquisition terminals on the associated equipment of the Manufacturing Execution System (MES). Real-time data acquisition methods and process parameter comparison and analysis methods are used to collect actual process operating parameters, design target parameters, and the deviation between the two. At the same time, production cycle time data is collected to construct production execution deviation data, denoted as Dm. Real-time monitoring of equipment status and user feedback behavior during product operation and maintenance is performed by installing vibration acceleration sensors, temperature sensors, current acquisition sensors, and speed encoders on the equipment. Flume log collection and Elasticsearch log parsing programs are deployed on the operation and maintenance platform. Multi-sensor data acquisition and log text parsing methods are used to collect equipment operating vibration amplitude, equipment temperature change, operating current fluctuation, and speed change data. At the same time, equipment operation log records, fault alarm records, and user satisfaction rating data are collected to construct operation and maintenance feedback data, denoted as Do. User input data Du, design parameter change data Dd, production execution deviation data Dm, and operation and maintenance feedback data Do are processed in a unified manner. A unified identifier label is added to the data from different sources by executing the data identification encoding method. The time-series synchronization of multi-source data is performed by the timestamp alignment method. The data expression differences are eliminated by the semantic normalization processing method to construct a unified dataset, denoted as Dall.

[0028] In this embodiment, by uniformly collecting and standardizing multi-source heterogeneous data such as user requirements, design changes, production execution, and operation and maintenance feedback, efficient integration and consistent expression of data are achieved throughout the entire lifecycle, providing a reliable data foundation for subsequent requirements analysis and modeling.

[0029] Example 3 Please see Figures 1 to 3 This embodiment, as explained in Embodiment 2, specifically involves extracting parameters such as the frequency of requirement changes, the degree of requirement conflict, the deviation from requirement implementation, the deviation from user feedback, and the duration of each phase. This includes the following steps: Extract the requirement identifier and requirement submission timestamp information from the user requirement input data Du, and extract the design parameter modification record and version iteration time information corresponding to the requirement identifier from the design parameter change data Dd. The number of change operations corresponding to the same requirement identifier is cumulatively counted using a preset time window as the statistical unit, and the total number of requirement items within the current time window is counted. The number of change operations and the total number of requirement items are proportionalized to obtain the requirement change frequency parameter, which represents the activity level of requirement changes per unit time, and is denoted as Fc. The textual requirement content and priority identification information are extracted from the user requirement input data Du, and the textual requirement content is semantically parsed to obtain the semantic expression result of the requirement. At the same time, the design parameter modification records in the design parameter change data Dd are combined to identify the overlap of different requirement items in terms of resource occupation and functional realization goals. Based on semantic consistency analysis and resource conflict judgment rules, requirement items with logical conflicts or resource conflicts are marked and their number is counted. Then, the total number of requirements in the same stage is counted, and the number of conflicting requirement items is proportionalized to the total number of requirements to obtain the requirement conflict degree parameter, denoted as Cd, which represents the degree of internal contradiction of the requirement. The actual process operation parameter data collected by industrial sensors is extracted from the production execution deviation data Dm, and the corresponding design target parameter data is extracted from the design parameter change data Dd. Each set of actual process operation parameters is compared with the design target parameters to obtain the deviation range between each parameter. The deviation ranges of all parameters are summarized to form the overall deviation level. The overall deviation level is used as the requirement realization deviation parameter, denoted as Id, to characterize the degree of deviation of the actual production execution process from the design target. Extract user satisfaction score data and user evaluation records within the corresponding time period from the operation and maintenance feedback data Do, and obtain the pre-set target satisfaction reference value. Perform difference analysis on the user satisfaction score data and the target satisfaction reference value to form a deviation result. Use the deviation result as a user feedback deviation parameter to characterize the degree of difference between the user's actual experience and the expected goal, denoted as Ub. The start and end timestamps corresponding to each lifecycle stage are extracted from user demand input data Du, design parameter change data Dd, production execution deviation data Dm, and operation and maintenance feedback data Do. The time difference calculation is performed on the start and end timestamps to obtain the actual duration of each stage. The actual duration is used as the stage duration parameter, which characterizes the length of the lifecycle stage operation cycle and is denoted as Ts. The maximum and minimum values ​​of each parameter in the historical data sample are obtained, and the demand change frequency parameter Fc, demand conflict degree parameter Cd, demand implementation deviation degree parameter Id, user feedback deviation parameter Ub and stage duration parameter Ts are normalized based on the interval scaling method. The processing results are mapped to the [0,1] interval to obtain normalized demand feature parameters, and a demand feature set is constructed based on the normalized demand feature parameters.

[0030] In this embodiment, by constructing multi-dimensional demand feature parameters and performing dimensionless processing, data from different sources and at different scales are made comparable and computable, thereby improving the stability of the precision in demand state characterization.

[0031] Example 4 Please see Figures 1 to 3 In the explanation of Example 3, this embodiment specifically describes the construction of the demand evolution index, which includes the following steps: Extract the following parameters from the demand feature set: demand change frequency (Fc), demand conflict degree (Cd), demand implementation deviation (Id), user feedback deviation (Ub), and stage duration (Ts). Input these parameters into the demand evolution modeling process. Perform weighted coupling calculations according to the historical regression weight coefficients corresponding to each parameter to obtain the demand evolution index, which characterizes the overall evolution status of demands across multiple stages of the product lifecycle. This index is denoted as... The formula is as follows: in, , , , and These represent the weighting coefficients, which are the weighting coefficients for the requirement change frequency parameter Fc, requirement conflict degree parameter Cd, requirement implementation deviation parameter Id, user feedback deviation parameter Ub, and phase duration parameter Ts, respectively. It should be noted that... Each weighting coefficient is set based on historical data regression results and business focus, and is used to balance the influence of different demand characteristic parameters on the demand evolution index. The degree of contribution, for example, in scenarios where requirements change frequently or user feedback is sensitive. or The value of [value] will be significantly higher than other weights to enhance the responsiveness to key influencing factors. The system employs a historical sample-driven adaptive weight update mechanism during demand evolution modeling. This mechanism involves analyzing the deviation between historical demand evolution results and actual execution effects to extract the contribution of each parameter to the overall evolution trend, and then dynamically correcting the weight coefficients accordingly, thereby achieving a more accurate assessment of the demand evolution index. Continuous optimization of the computational model, while ensuring the synergistic effect of various feature parameters, improves the model's accuracy and adaptability to complex and changing environments.

[0032] Demand Evolution Index Essentially, it is a comprehensive representation quantity that combines multiple factors with weighted summaries. Its physical meaning lies in mapping several key driving factors influencing demand changes (change frequency, conflict level, implementation deviation, user feedback deviation, and time factor) onto a unified metric space. Through linear weighted coupling, it reflects the comprehensive strength of the system's "demand disturbance energy." Each parameter can be considered a "disturbance source" of different dimensions, and the weighting coefficients reflect the proportion of each disturbance source's contribution to the overall system state. Therefore, the demand evolution index... It can be understood as the overall deviation or evolutionary activity of the demand system under the influence of multiple driving forces. The larger the value, the more obvious the deviation of the demand system from the initial stable state.

[0033] In this embodiment, a demand evolution index is formed by weighted coupling calculation of key demand characteristic parameters, thereby achieving a quantitative description of the complex demand change process and effectively improving the ability to identify demand change trends and the level of comprehensive evaluation.

[0034] Example 5 Please see Figures 1 to 3 In the explanation of Example 4, this embodiment specifically describes the triggering of the demand adjustment strategy, which includes the following steps: Demand Evolution Index A comparative analysis is performed between the pre-set primary evolution threshold (T1) and the secondary risk threshold (T2), and the demand evolution index. When the demand evolution index is ≤ the first-level evolution threshold T1, the current demand evolution is determined to be in a stable and controllable state and meets the qualification requirements. When the first-level evolution threshold T1 is greater than or equal to the second-level risk threshold T2, the current demand evolution is determined to be in an unbalanced state and there is a risk of local deviation. When the demand evolution index... >When the secondary risk threshold T2 is reached, it is determined that the current demand evolution is in an abnormally out-of-control state and there is a risk of systemic demand mismatch; When the demand evolution index When in an unbalanced state, a 10% to 20% freeze ratio is applied to change requests corresponding to the demand change frequency parameter, and a priority reordering adjustment of no less than 30% is applied to conflicting requests corresponding to the demand conflict degree parameter. Simultaneously, a 5% to 10% reverse correction adjustment is implemented to the production execution parameters corresponding to the demand realization deviation parameter to reduce deviation accumulation. When the demand evolution index... When in an abnormal out-of-control state, at least 50% of the change requirements corresponding to the demand change frequency parameter will be forcibly frozen, and at least 60% of the conflicting requirements corresponding to the demand conflict degree parameter will be prioritized and restructured. At the same time, 15% to 30% of the production execution parameters corresponding to the demand realization deviation degree parameter will be forcibly reversed and adjusted, and cross-stage demand backtracking analysis will be triggered simultaneously to eliminate systemic risks. The hierarchical demand adjustment strategy is transformed into a set of control instructions that includes parameters such as the demand freeze ratio, priority adjustment weight, and production parameter correction magnitude. A demand adjustment execution dataset is then constructed and distributed to the demand management, design adjustment, and production execution stages according to the product lifecycle stage mapping relationship, so as to achieve closed-loop intervention and control of the multi-stage demand evolution process.

[0035] The goal of calibrating the primary evolution threshold T1 and the secondary risk threshold T2 is to determine the boundary intervals of the demand evolution index that can effectively distinguish between "stable demand evolution state," "unbalanced demand evolution state," and "abnormally out-of-control demand state." First, a standard data database of demand evolution states is constructed. This database is formed by continuously collecting demand-related data at each stage of the product lifecycle, with a sample size of no less than 800 sets, covering various typical states such as stable demand operation, partial imbalance, and severe out-of-control demand. Each sample set includes a demand change frequency parameter Fc, a demand conflict degree parameter Cd, a demand implementation deviation parameter Id, a user feedback deviation parameter Ub, and a stage duration parameter Ts. Multiple demand analysis experts and project managers, combining project execution results and user satisfaction evaluations, use a "gold standard" to label these parameters, classifying them into "stable state," "unbalanced state," and "abnormally out-of-control state." Subsequently, all samples were uniformly calculated using the demand evolution index formula, and the distribution of the three types of state data was statistically analyzed to construct a probability density distribution model. Based on this, a multi-class ROC curve analysis method was employed, and two optimal dividing points were determined using the principle of maximizing the Youden index. These were designated as the primary evolution threshold T1 and the secondary risk threshold T2, respectively. Furthermore, the thresholds were adjusted based on engineering experience; for example, when the demand evolution index exceeds a certain level, demand conflict and execution deviation increase significantly. Finally, T1 and T2 were selected as the dividing points between stability and imbalance, and between imbalance and out of control, respectively, thereby achieving hierarchical identification and precise control of demand states.

[0036] In this embodiment, a tiered demand adjustment strategy is triggered based on a threshold range to achieve differentiated intervention and control of demand states at different risk levels, effectively reducing the risk of demand runaway and improving the flexibility and accuracy of system response.

[0037] Example 6 Please see Figures 1 to 3In the explanation of Example 5, this embodiment specifically describes the construction of the demand consistency index and the demand stability index, including the following steps: Based on the actual process operating parameters in the production execution deviation data Dm and the design parameter modification records in the design parameter change data Dd, a parameter identifier alignment method is used to match the design parameters with the actual process operating parameters. The design target parameter value corresponding to the i-th parameter is then extracted according to a unified parameter number order, denoted as […]. And the actual execution parameter value corresponding to the i-th parameter, denoted as Meanwhile, when performance metrics or outcome evaluation data exist in the operation and maintenance feedback data (Do), a data mapping method is used to convert the performance metrics into corresponding parameter dimension data and supplement them to the actual execution parameter values. middle; Achievement value of design phase objectives Compared with actual execution results A step-by-step matching process is performed, and the absolute value of the difference between each matching item is calculated. After averaging all the absolute values ​​of the differences, inverse normalization is performed to obtain the design consistency index, which characterizes the degree of consistency between the design objective and the actual execution result. This index is denoted as [index not provided]. The formula is as follows: Where n represents the total number of parameters calculated; Design Consistency Index Essentially, it's a normalized measure of the error between the "target state" and the "actual state," its physical meaning stemming from error analysis theory. By statistically averaging the absolute values ​​of the differences between the actual and design target values ​​of each parameter, the overall system deviation level can be obtained. Then, through inverse mapping (1 - average error), the "error quantity" is converted into a "consistency measure." Therefore, the design consistency index... It can be understood as the "degree of approximation" of the system output to the design input. The closer its value is to 1, the closer the system execution is to the ideal design state, reflecting the accuracy and consistency level of the system.

[0038] Based on the demand evolution index The output results at each stage of the product lifecycle are obtained by using time series extraction methods to acquire demand evolution index sequence data corresponding to multiple stages. After calculating the overall mean of the data for each stage in the sequence, mean centering is performed on the data for each stage to obtain the deviation value. Then, all deviation values ​​are statistically processed using the cumulative square root method to obtain the standard deviation of the demand evolution index, denoted as . ; Standard deviation of the demand evolution index A normalization transformation is performed using a reverse mapping method to obtain a demand stability index, denoted as [index not provided in the original text]. The formula is as follows: Demand stability index Built upon the principles of time series volatility analysis, its core physical meaning is a measure of the "intensity of volatility" in the demand evolution process. This is achieved by calculating the standard deviation of the demand evolution index across multiple stages. The standard deviation can reflect the system's dispersion or fluctuation amplitude over time; a larger standard deviation indicates more severe fluctuations in the system's state. Further, through inverse mapping... This transforms volatility into a stability index, aligning it with the intuitive physical principle that "lower volatility equals higher stability." Therefore, the demand stability index... Essentially, it is a quantitative representation of the system's operational stability or its ability to resist disturbances.

[0039] In this embodiment, by constructing a design consistency index and a demand stability index, a two-dimensional quantitative assessment of system execution deviation and operational fluctuations is achieved, thereby improving the comprehensive judgment ability on system operational quality and stability.

[0040] Example 7 Please see Figures 1 to 3 This embodiment, as explained in Embodiment Six, specifically outlines the construction of a dual-dimensional evaluation system for consistency and stability, along with corresponding optimization strategies, including the following steps: Design Consistency Index The design consistency threshold is compared with the preset threshold, denoted as T3, and the demand stability index is also considered. The design consistency index is compared with the preset operational stability threshold, denoted as T4; When the design consistency threshold T3 is reached, the current system is deemed to meet the design consistency requirements at the parameter execution level, and the system design execution consistency status is deemed qualified; when the design consistency index is reached... When the design consistency threshold T3 is reached, the system design execution consistency is deemed unqualified, indicating a risk of target deviation and accumulated execution error; when the demand stability index... When the system stability index is ≥ the operational stability threshold T4, the system is considered to be operating stably; when the demand stability index is ≥ the operational stability threshold T4, the system is considered to be operating stably. If the system is less than the stability judgment threshold T4, it is determined that the system is unstable and there is a risk of scheduling fluctuations and resource mismatch. When the design consistency index <Design consistency judgment threshold T3 and demand stability index When the system's operational stability threshold T4 is reached, it is determined to be in a doubly unqualified state, triggering a forced collaborative optimization strategy. This involves a comprehensive backtracking correction of the design target parameters, a 20%–30% regression correction of key execution parameters based on the deviation ratio, a 15%–25% reallocation of resource priority weights, and a 10%–20% compression or extension adjustment of the production scheduling cycle time to eliminate the risk of coupling between systemic deviations and fluctuations. When the design consistency index... <Design consistency judgment threshold T3 and demand stability index When the value is ≥ the operational stability threshold T4, it is determined to be in a state of unqualified consistency but qualified stability, triggering a targeted consistency correction strategy. This involves performing a 10%–20% fine-grained compensation adjustment on the parameters representing the source of the deviation, while maintaining the current scheduling rhythm to avoid introducing new fluctuation risks. When the design consistency index... ≥ Design consistency threshold T3 and demand stability index When the system reaches the stability threshold T4, it is determined to be in a state of consistency qualification but stability qualification, triggering a stability enhancement strategy. This involves smoothing the resource scheduling rhythm by 10% to 25%, and dynamically rebalancing the resource allocation ratio of high-fluctuation links by 10% to 20% to improve the continuity of system operation. The system also records the parameter correction, resource reallocation and scheduling adjustment results to construct a collaborative optimization execution result dataset.

[0041] The goal of calibrating the design consistency judgment threshold T3 is to determine a critical value for the design consistency index that can effectively distinguish between "design execution consistency" and "design execution deviation." First, a design execution consistency assessment database is constructed by collecting design target parameters and actual execution parameters from historical projects to form a sample set of no less than 600 groups, covering different situations such as high consistency, moderate deviation, and severe deviation. Design engineers and quality engineers then label the consistency status based on test results and performance compliance. Next, the design consistency index is calculated for each sample, and distribution analysis is performed on the consistent and inconsistent samples to construct a probability density function. Further, ROC curve analysis is used to select the boundary point with the largest Youden index as the initial threshold, which is then adjusted based on engineering tolerance requirements; for example, when the parameter deviation exceeds the allowable range of the process, it is judged as inconsistent. Finally, the design consistency judgment threshold T3 is determined to accurately identify the risk of design execution deviation.

[0042] The goal of calibrating the operational stability judgment threshold T4 is to determine a critical value for the demand stability index that can distinguish between "stable operation" and "fluctuating operation." First, an operational stability sample database is constructed by extracting stage sequence data of the demand evolution index to form time series samples, with a sample size of no less than 700 groups, covering various states such as stable operation, slight fluctuations, and severe fluctuations. System maintenance personnel then label the states based on operational continuity, scheduling fluctuations, and resource matching. Subsequently, the demand stability index is calculated for each sample, and statistical distribution analysis is performed on stable and unstable samples. Furthermore, the optimal cutoff point is determined using ROC curves and the Youden index method, while adjustments are made based on system scheduling experience; for example, fluctuations exceeding a certain threshold will significantly affect production continuity. Finally, the operational stability judgment threshold T4 is determined to effectively identify the risk of system operational fluctuations.

[0043] In this embodiment, by establishing a consistency and stability combination judgment mechanism and matching differentiated collaborative optimization strategies, targeted correction and control of different abnormal states can be achieved, effectively suppressing the accumulation of deviations and the risk of system fluctuations.

[0044] Example 8 Please see Figures 1 to 3 The following is an explanation of Example 7, specifically the construction of demand deviation tracing and impact path analysis, which includes the following steps: Based on production execution deviation data Dm, design parameter change data Dd, demand adjustment execution dataset, and collaborative optimization execution result dataset, this paper uses a phase division and data association mapping method to identify and divide each stage of the product lifecycle. It extracts the actual execution deviation data corresponding to each stage from the production execution deviation data Dm, and combines this with the design target parameter data from the design parameter change data Dd. The difference between the actual deviation and the design target deviation for each stage is calculated and normalized. Finally, the total deviation across the entire lifecycle is proportionally allocated to obtain the deviation contribution value for the k-th stage, denoted as [missing value]. Based on the demand adjustment execution dataset and the collaborative optimization execution result dataset, a multiple regression analysis method is used to fit and calculate the degree of influence of demand adjustment behavior and collaborative optimization adjustment behavior at each stage on the overall deviation propagation, obtaining the stage influence weight corresponding to the k-th stage, denoted as . The weighted cumulative calculation method is used to calculate the contribution value of deviations at each stage. Weight of stage impact We perform weighted summation to construct the demand deviation propagation index, denoted as... The calculation formula is as follows: Where m represents the total number of stages in the product lifecycle.

[0045] Demand Deviation Propagation Index Derived from the principles of "multi-stage coupling propagation" and "weighted superposition effect," its physical meaning lies in describing the accumulation and amplification process of deviations across different stages of the product lifecycle. This represents the proportion of each stage's contribution to the overall deviation, and can be regarded as the "source strength" of the local deviation. This represents the amplification or suppression effect of this stage on deviation propagation, equivalent to the "propagation gain coefficient." By weighted summing the deviation contribution and propagation impact of each stage, the overall deviation propagation intensity of the system can be obtained. Therefore, the demand deviation propagation index... It can be understood as the comprehensive propagation energy of the deviation in a multi-stage system. The larger the value, the higher the degree of diffusion of the deviation in the system and the more significant the risk.

[0046] In this embodiment, by constructing a demand deviation propagation index and analyzing the contribution and impact weight of deviations at each stage, the source and propagation path of deviations are quantitatively identified, providing a basis for accurately locating key influencing links and optimizing decisions.

[0047] Example 9 Please see Figures 1 to 3 In the explanation of Example 8, this embodiment specifically describes the construction of an execution feedback and closed-loop optimization update mechanism, which includes the following steps: Demand deviation propagation index Compared with the preset optimization convergence judgment threshold, denoted as T5, when the demand deviation propagation index... When the optimization convergence threshold T5 is reached, the system deviation propagation is considered to be in a convergent state, the system optimization state is deemed satisfactory and there is no significant propagation risk, and the current demand adjustment strategy and collaborative optimization strategy are maintained; when the demand deviation propagation index is... > When optimizing the convergence threshold T5, the system is deemed to be in an unqualified optimization state, posing a risk of bias propagation and path amplification. This triggers an iterative optimization strategy, adjusting the stage influence weights corresponding to high-impact stages. Implement a 5%–15% suppression adjustment, and simultaneously implement enhanced interventions of demand adjustment strategy and collaborative optimization strategy to reduce the intensity of deviation propagation; Adjusting the phase impact weights after strategy adjustment Feedback to the demand deviation propagation index The calculation process is carried out, and the demand adjustment execution dataset and the collaborative optimization execution result dataset are updated synchronously to form a full life cycle closed-loop optimization control mechanism driven by the demand deviation propagation index.

[0048] The calibration goal of optimizing the convergence threshold T5 is to determine the critical value of the required deviation propagation index that can distinguish between the "deviation propagation convergence state" and the "deviation diffusion state". First, a deviation propagation analysis database is constructed by collecting deviation contribution values ​​at each stage during multi-cycle operation. and stage influence weight The data forms a sample set with no fewer than 600 samples, covering various states such as gradual convergence, slow fluctuation, and continuous propagation of deviations. Quality analysts label these states based on the final product quality results and deviation trends. Subsequently, the demand deviation propagation index is calculated for each sample, and distribution statistics and probability density analysis are performed on converged and non-converged samples. Furthermore, ROC curves and the Youden index are used to determine the optimal threshold point, which is then adjusted based on engineering experience; for example, exceeding a certain level of deviation propagation will lead to systemic quality risks. Finally, an optimization convergence judgment threshold T5 is determined to determine whether the system has reached a stable optimization state and to guide the execution of subsequent closed-loop control strategies.

[0049] In this embodiment, a threshold determination and feedback update mechanism based on the deviation propagation index is used to achieve continuous evaluation and dynamic adjustment of the strategy execution effect, forming an adaptive closed-loop optimization process, thereby improving the stability and convergence capability of the system in the long term.

[0050] Example 10 Please refer to the full lifecycle demand intelligent analysis system. Figures 1 to 3 Specifically, including: The full lifecycle requirement data fusion and processing module is used to parse and obtain multi-source requirement data streams at each stage of the product lifecycle, extract requirement semantic data, design execution data, production deviation data and operation and maintenance feedback data, and perform unified identification encoding, timestamp alignment and semantic normalization processing on various types of data to build a unified dataset. The standardized requirement feature parameter construction module is used to extract requirement change frequency parameters, requirement conflict degree parameters, requirement implementation deviation degree parameters, user feedback deviation parameters, and phase duration parameters based on a unified dataset, and perform dimensionless processing to establish a requirement feature set. The demand evolution assessment and control module is used to calculate and construct the demand evolution index, compare the demand evolution index with the first-level evolution threshold and the second-level risk threshold, determine whether the demand is stable, unbalanced or abnormally out of control, and execute a graded demand adjustment strategy accordingly, including demand freezing, priority adjustment and production parameter correction, and form a demand adjustment execution dataset. The consistency and stability analysis and optimization module is used to calculate the design consistency index and, based on the stage sequence data of the demand evolution index, calculate the standard deviation and perform inverse mapping to obtain the demand stability index. The design consistency index is compared with the design consistency judgment threshold and the demand stability index is compared with the operational stability judgment threshold to determine the consistency and stability status of the system. Differentiated collaborative optimization strategies are triggered according to different combinations of states, including backtracking correction, targeted compensation or stability enhancement adjustment. At the same time, the adjustment results are recorded and a collaborative optimization execution result dataset is constructed. The consistency and stability analysis and optimization module is used to calculate the demand deviation propagation index and conduct quantitative analysis of the sources and propagation paths of demand deviations. It compares the demand deviation propagation index with the optimization convergence judgment threshold to determine whether the system is in a deviation convergence state. If it does not converge, it iteratively strengthens and adjusts the weights of high-impact stages and related adjustment and optimization strategies, and feeds the updated results back to the preceding calculation and control process to form a closed-loop optimization control mechanism for the entire life cycle.

[0051] In this embodiment, the intelligent analysis system for full lifecycle requirements is constructed in a modular manner, which enables the decoupling and collaborative operation of each functional unit. This is beneficial for the engineering implementation, functional expansion and practical deployment of the system, and improves the maintainability and application efficiency of the overall system.

[0052] It should be noted that all calculation formulas in this application employ regression analysis, including but not limited to machine learning algorithms, to deeply analyze the collected parameters and identify their natural trends and interrelationships. Specialized software, such as Python's Scikit-learn library or the R language, is used to automatically generate mathematical models that match the data. Then, cross-validation and other methods are used to objectively evaluate the model performance, and continuous feedback and optimization are combined to ensure that the created formulas truly reflect the inherent laws of the data, thereby guaranteeing their effectiveness and accuracy. In all calculation formulas in this application, the parameters in each formula undergo dimensionless processing within a consistent range to ensure that different physical quantities are compared on the same scale; dimensionless processing techniques include, but are not limited to, min-max-normalization and Z-score standardization. The algorithm of this invention is implemented as a Python script. Before executing the core logic, the program first executes a data loading module (e.g., using the widely used pandas library in Python) configured to read the aforementioned spreadsheet file and load its contents into the program's working memory (e.g., a DataFrame data structure). Subsequent algorithm steps will directly query and retrieve the required configuration parameters from this in-memory data structure.

[0053] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A full lifecycle demand intelligent analysis method, characterized by: The specific steps include: S1. Parse and obtain multi-source demand data streams at each stage of the product lifecycle, extract demand semantic data, design execution data, production deviation data and operation and maintenance feedback data, and perform unified identification encoding, timestamp alignment and semantic normalization processing on various types of data to build a unified dataset. S2. Extract the frequency of requirement changes, degree of requirement conflict, degree of deviation of requirement implementation, deviation of user feedback, and duration of each phase based on the unified dataset, and perform dimensionless processing to establish a requirement feature set. S3. Based on the demand feature set, construct the demand evolution index, and filter according to the demand evolution index to determine whether the demand is in a stable, unbalanced or abnormally out-of-control state. Based on this, execute the hierarchical demand adjustment strategy, including demand freezing, priority adjustment and production correction, and form a demand adjustment execution dataset. S4. Based on the demand adjustment execution dataset, calculate the design consistency index. At the same time, calculate the standard deviation based on the stage sequence data of the demand evolution index and perform inverse mapping to obtain the demand stability index. Then, screen and determine the system consistency and stability status. Trigger differentiated collaborative optimization strategies according to different combination states, including backtracking correction, targeted compensation or stability enhancement adjustment. Record the adjustment results and construct a collaborative optimization execution result dataset. S5. Based on the collaborative optimization execution result dataset, calculate the demand deviation propagation index to conduct a quantitative analysis of the sources and propagation paths of demand deviations; determine whether the deviation is in a convergence state based on the demand deviation propagation index.

2. The intelligent lifecycle demand analysis method according to claim 1, characterized in that: A unified dataset includes the following steps: Real-time monitoring of user interaction behavior at each stage of the product lifecycle; collection of textual requirements submitted by users, timestamps of requirement submission, and priority identifiers of requirements; and construction of user requirement input data. Real-time monitoring of parameter evolution behavior during product design process; collection of design version iteration count, design parameter modification records, and time information corresponding to each change; and construction of design parameter change data. The execution status of the production process is monitored in real time, and actual process operation parameters, design target parameters and the deviation between the two are collected. At the same time, production cycle time data is collected to construct production execution deviation data. Real-time monitoring of equipment status and user feedback behavior during product operation and maintenance; collection of equipment vibration amplitude, equipment temperature change, operating current fluctuation and speed change data; collection of equipment operation log records, fault alarm records and user satisfaction score data; and construction of operation and maintenance feedback data. User input data, design parameter change data, production execution deviation data, and operation and maintenance feedback data are processed in a unified manner. A unified identifier label is added to data from different sources through execution data identification and encoding methods. A timestamp alignment method is used to synchronize the time sequence of multi-source data. Semantic normalization processing is used to eliminate data expression differences and construct a unified dataset.

3. The intelligent lifecycle demand analysis method according to claim 2, characterized in that: Extract the frequency of requirement changes, the degree of requirement conflict, the deviation from requirement implementation, the deviation from user feedback, and the duration of each phase, including the following steps: Extract requirement identifiers and requirement submission timestamps from user requirement input data, and extract design parameter modification records and version iteration time information corresponding to the requirement identifiers from design parameter change data. Using a preset time window as the statistical unit, accumulate the number of change operations corresponding to the same requirement identifier and count the total number of requirement items within the current time window. Proportionate the number of change operations to the total number of requirement items to obtain the requirement change frequency. The textual requirement content and priority identification information are extracted from the user requirement input data, and the textual requirement content is semantically parsed to obtain the semantic expression result of the requirement. At the same time, the design parameter modification records in the design parameter change data are combined to identify the overlap of different requirement items in terms of resource occupation and functional realization goals. Based on semantic consistency analysis and resource conflict judgment rules, requirement items with logical conflicts or resource conflicts are marked and their number is counted. Then, the total number of requirements in the same stage is counted, and the number of conflicting requirement items is proportionalized to the total number of requirements to obtain the requirement conflict degree. The actual process operation parameter data collected by industrial sensors is extracted from the production execution deviation data, and the corresponding design target parameter data is extracted from the design parameter change data. Each set of actual process operation parameters and design target parameters are compared item by item to obtain the deviation range between each parameter. The deviation ranges of all parameters are summarized to form the overall deviation level, and the overall deviation level is used as the deviation degree of requirement realization. Extract user satisfaction rating data and user evaluation records within the corresponding time period from the operation and maintenance feedback data, obtain the pre-set target satisfaction reference value, perform difference analysis on the user satisfaction rating data and the target satisfaction reference value to form the deviation result, and use the deviation result as the user feedback deviation. The start and end timestamps corresponding to each lifecycle stage are extracted from user demand input data, design parameter change data, production execution deviation data and operation and maintenance feedback data. The time difference between the start and end timestamps is calculated to obtain the actual duration of each stage, and the actual duration is used as the stage duration. The maximum and minimum values ​​of each parameter in the historical data sample are obtained, and the frequency of requirement change, degree of requirement conflict, degree of requirement implementation deviation, user feedback deviation and stage duration are normalized based on the interval scaling method. The processing results are mapped to the [0,1] interval to obtain normalized requirement features, and a requirement feature set is constructed based on the normalized requirement features.

4. The intelligent lifecycle demand analysis method according to claim 3, characterized in that: Constructing a demand evolution index includes the following steps: Extract the frequency of demand changes, degree of demand conflict, degree of demand implementation deviation, user feedback bias, and stage duration from the demand feature set, input them into the demand evolution modeling process, and perform weighted coupling calculation according to the historical regression weight coefficients corresponding to each parameter to obtain the demand evolution index.

5. The intelligent lifecycle demand analysis method according to claim 4, characterized in that: Triggering demand adjustment strategies includes the following steps: By comparing the demand evolution index with the preset first-level evolution threshold and second-level risk threshold, the demand is identified as being in a stable, unbalanced, or abnormally out-of-control state. The tiered demand adjustment strategy is transformed into a set of control instructions that includes the demand freeze ratio, priority adjustment weight, and production parameter correction range, and a demand adjustment execution dataset is constructed; and distributed to the demand management, design adjustment, and production execution stages according to the product life cycle stage mapping relationship.

6. The intelligent lifecycle demand analysis method according to claim 5, characterized in that: Constructing the demand consistency index and demand stability index includes the following steps: Based on the actual process operation parameters in the production execution deviation data and the design parameter modification records in the design parameter change data, the parameter identification alignment method is used to match the design parameters with the actual process operation parameters. The design target parameter value corresponding to the i-th parameter and the actual execution parameter value corresponding to the i-th parameter are extracted according to the unified parameter number order. At the same time, when there are performance indicators or result evaluation data in the operation and maintenance feedback data, the data mapping method is used to convert the performance indicators into corresponding parameter dimension data and supplement them into the actual execution parameter values. The design phase target values ​​and actual execution results are matched item by item, and the absolute value of the difference between each matching item is calculated. After the absolute value of all differences is averaged, the reverse normalization calculation is performed to obtain the design consistency index, which is used to characterize the degree of consistency between the design target and the actual execution results. Based on the output results of the demand evolution index in each stage of the product life cycle, the time series extraction method is used to obtain the demand evolution index sequence data corresponding to multiple stages. The mean calculation method is used to obtain the overall mean of the data in each stage of the sequence. Then, mean centering is performed on the data of each stage to obtain the deviation value. Finally, the cumulative square and square root calculation method is used to perform statistical processing on all deviation values ​​to obtain the standard deviation of the demand evolution index. The standard deviation of the demand evolution index is normalized using a reverse mapping method to obtain a demand stability index that characterizes the stability of fluctuations in the demand evolution process.

7. The intelligent lifecycle demand analysis method according to claim 6, characterized in that: Constructing a dual-dimensional evaluation framework for consistency and stability, along with corresponding optimization strategies, includes the following steps: The design consistency index is compared with the preset design consistency judgment threshold, and the demand stability index is compared with the preset operation stability judgment threshold; thus, scheduling fluctuation risk and resource mismatch risk are screened out. In response to the dual risks, a mandatory collaborative optimization strategy is triggered; In response to the risk of resource mismatch, a targeted consistency correction strategy is triggered; To address scheduling fluctuation risks, a stability enhancement strategy is triggered; and the results of parameter correction, resource reallocation, and scheduling adjustments are recorded to construct a dataset of collaborative optimization execution results.

8. The intelligent lifecycle demand analysis method according to claim 7, characterized in that: Constructing a demand deviation source tracing and impact path analysis includes the following steps: Based on production execution deviation data, design parameter change data, demand adjustment execution dataset, and collaborative optimization execution result dataset, this study uses a phase division and data association mapping method to identify and divide each stage of the product lifecycle. Actual execution deviation data for each stage is extracted from the production execution deviation data. Simultaneously, design target parameter data from the design parameter change data is combined to calculate and normalize the difference between the actual deviation and the design target deviation for each stage. The total deviation across the entire lifecycle is then proportionally allocated to obtain the deviation contribution value for the k-th stage. Based on the demand adjustment execution dataset and the collaborative optimization execution result dataset, a multiple regression analysis method is used to fit and calculate the impact of demand adjustment behavior and collaborative optimization adjustment behavior on the overall deviation propagation, obtaining the stage influence weight corresponding to the k-th stage. Finally, a weighted cumulative calculation method is used to sum the deviation contribution value and stage influence weight for each stage to construct a demand deviation propagation index.

9. The intelligent lifecycle demand analysis method according to claim 8, characterized in that: Constructing an execution feedback and closed-loop optimization and update mechanism includes the following steps: By comparing the demand deviation propagation index with the preset optimization convergence threshold, the risks of deviation propagation and diffusion, as well as path amplification risks, are identified, triggering iterative optimization strategies. The stage impact weights after strategy adjustment are fed back into the calculation process of demand deviation propagation index, and the demand adjustment execution dataset and collaborative optimization execution result dataset are updated synchronously.

10. A full lifecycle demand intelligent analysis system, applied to the full lifecycle demand intelligent analysis method according to any one of claims 1 to 9, characterized in that, include: The full lifecycle requirement data fusion and processing module is used to parse and obtain multi-source requirement data streams at each stage of the product lifecycle, extract requirement semantic data, design execution data, production deviation data and operation and maintenance feedback data, and perform unified identification encoding, timestamp alignment and semantic normalization processing on various types of data to build a unified dataset. The standardized requirement feature parameter construction module is used to extract requirement change frequency, requirement conflict degree, requirement implementation deviation degree, user feedback deviation and stage duration based on a unified dataset, and perform dimensionless processing to establish a requirement feature set; The demand evolution assessment and control module constructs a demand evolution index based on the demand feature set, and filters according to the demand evolution index to determine whether the demand is in a stable, unbalanced or abnormally out-of-control state. Based on this, it executes a graded demand adjustment strategy, including demand freezing, priority adjustment and production correction, and forms a demand adjustment execution dataset. The consistency and stability analysis and optimization module is used to adjust the execution dataset according to demand, calculate the design consistency index, calculate the standard deviation based on the stage sequence data of the demand evolution index and perform inverse mapping to obtain the demand stability index, and then screen and determine the consistency and stability status of the system. It also triggers differentiated collaborative optimization strategies based on different combinations of states, including backtracking correction, targeted compensation or stability enhancement adjustment, while recording the adjustment results and constructing a collaborative optimization execution result dataset. The consistency and stability analysis and optimization module calculates the demand deviation propagation index based on the collaborative optimization execution result dataset, and performs quantitative analysis on the sources and propagation paths of demand deviations; it also determines whether the deviation is in a convergence state based on the demand deviation propagation index.

Images